Photovoltaic cells convert light into electric current. One of the most important measures of a photovoltaic cell is its efficiency in converting light energy into electrical energy. Although photovoltaic cells can be fabricated from a variety of semiconductor materials, silicon is generally used because it is readily available at reasonable cost, and because it has a suitable balance of electrical, physical, and chemical properties for use in fabricating photovoltaic cells.
In a known procedure for the manufacture of photovoltaic cells, silicon feedstock is mixed with a material (or dopant) for inducing either a positive or negative conductivity type, melted, and then crystallized by either pulling crystallized silicon out of a melt zone into ingots of monocrystalline silicon (via the Czochralski (CZ) or float zone (FZ) methods), or cast into blocks or “bricks” of multi-crystalline silicon or polycrystalline silicon, depending on the grain size of the individual silicon grains. In the procedure described above, the ingots or blocks are cut into thin substrates, also referred to as wafers, by known slicing or sawing methods. These wafers may then be processed into photovoltaic cells.
Monocrystalline silicon for use in the manufacture of photovoltaic cells is generally produced by the CZ or FZ methods, both being processes in which a cylindrically shaped boule of crystalline silicon is produced. For a CZ process, the boule is slowly pulled out of a pool of molten silicon. For a FZ process, solid material is fed through a melting zone and re-solidified on the other side of the melting zone. A boule of monocrystalline silicon, manufactured in these ways, contains a radial distribution of impurities and defects, such as rings of oxygen-induced stacking faults (OSF) and “swirl” defects of interstitial or vacancy clusters. Even with the presence of these impurities and defects, monocrystalline silicon is generally a preferred source of silicon for producing photovoltaic cells, because it can be used to produce high efficiency solar cells. Monocrystalline silicon is, however, more expensive to produce than conventional multi-crystalline silicon, using known techniques such as those described above.
Conventional multi-crystalline silicon for use in the manufacture of photovoltaic cells is generally produced by a casting process. Casting processes for preparing conventional multi-crystalline silicon are known in the art of photovoltaic technology. Briefly, in such processes, molten silicon is contained in a crucible, such as a quartz crucible, and is cooled in a controlled manner to permit the crystallization of the silicon contained therein. The block of multi-crystalline silicon that results is generally cut into bricks having a cross-section that is the same as or close to the size of the wafer to be used for manufacturing a photovoltaic cell, and the bricks are sawed or otherwise cut into such wafers. The multi-crystalline silicon produced in such manner is an agglomeration of crystal grains where, within the wafers made therefrom, the orientation of the grains relative to one another is effectively random.
The random orientation of grains, in either conventional multi-crystalline or poly-crystalline silicon, makes it difficult to texture the surface of a resulting wafer. Texturing is used to improve efficiency of a photovoltaic cell, by reducing light reflection and improving light energy absorption through the surface a cell. Additionally, “kinks” that form in the boundaries between the grains of conventional multi-crystalline silicon tend to nucleate structural defects in the form of clusters or lines of dislocations. These dislocations, and the impurities they tend to attract, are believed to cause a fast recombination of electrical charge carriers in a functioning photovoltaic cell made from conventional multi-crystalline silicon. This can cause a decrease in the efficiency of the cell. Photovoltaic cells made from such multi-crystalline silicon generally have lower efficiency compared to equivalent photovoltaic cells made from monocrystalline silicon, even considering the radial distribution of defects present in monocrystalline silicon produced by known techniques. However, because of the relative simplicity and lower costs for manufacturing conventional multi-crystalline silicon, as well as effective defect passivation in cell processing, multi-crystalline silicon is a more widely used form of silicon for manufacturing photovoltaic cells.
Some previous casting techniques involved using a “cold-wall” crucible for crystal growth. The term “cold-wall” refers to the fact that induction coils present on or in the walls of the crucible are water cooled, and may also be slotted, thus generally remaining below 100.degree. C. The crucible walls may be situated in close proximity between the coils and the feedstock. The material of the crucible walls is not particularly thermally insulating, and can therefore remain in thermal equilibrium with the cooled coils. The heating of the silicon is therefore not predicated on radiation from the crucible walls, because inductive heating of the silicon in the crucible means that the silicon is heated directly by current induced to flow therein. In this way, the walls of the crucible remain below the melting temperature of the silicon, and are considered “cold,” relative to the molten silicon. During solidification of the inductively heated molten silicon, these cold walls of the crucible act as a heat sink. The ingot cools quickly, determined by radiation to the cold walls. Therefore, an initial solidification front quickly becomes substantially curved, with crystal nucleation occurring at the ingot sides and growing diagonally towards the ingot center, disrupting any attempt at maintaining a vertical and geometrically ordered seeding process or a substantially flat solidification front.
In view of the foregoing, there is a need for an improved form of silicon that can be used to manufacture photovoltaic cells. There is also a need for silicon that can be manufactured in a process that is faster and less expensive than the processes that have been heretofore used to produce monocrystalline silicon. The present invention provides such silicon and such processes.